The proposed study will provide information needed to improve the quality of care and life experienced by nursing home residents with mental status problems and behavioral disturbances. In particular, we intend to provide data and analyses that have implications for care and practice, for planning mental health services to the institutionalized elderly, and for development of public policies aimed at improving the quality of institutional care. The specific aims include; 1) describing the prevalence of certain mental status characteristics (e.g.., or orientation, memory) and disordered behaviors (e.g. wandering, yelling and verbal abuse, combativeness) as well as " clusters" of behaviors among nursing home residents; 2) describing institutional practices, such as use of psychotropics, physical restraints, and specialized services; 3) identifying factors (e.g., staffing levels) associated with differences in the use of various interventions; and 4) identifying characteristics and practices of "model" facilities having low use of chemical and physical restraints relative to the numbers of residents with behavioral disturbances. To accomplish these aims, we propose secondary analysis of six data sets that contain resident-level and facility-level information on 189,000 residents in 1,965 nursing homes in three states. We also propose primary data collection in six model facilities, concentrating on such features as staff attitudes and training, mental health services, use of volunteers, and environmental modifications that may facilitate management of residents without high use of physical and chemical restraints. The analytic techniques used to accomplish the first three aims include cross-tabulations examining the prevalence of disordered behaviors and of various interventions and logistic regressions to model the likelihood of a resident's being administered psychotropics and/or physical restraints as a result of facility characteristics (e.g. staffing levels), and resident characteristics (e.g. gender, payor, social resources, disordered behaviors). Describing"model" facilities involves on-site visits and primary data collection and case study observation and reporting. The project staff are a multidisciplinary team or researchers from RT1 as well AS consultants from Duke University's Aging Center, the University of Wisconsin, and the School of Pharmacy at the University of North Carolina.